Document Sample
DRYING (DOC) Powered By Docstoc
					Wheaton, F.W., Lawson T. B. 1985. Processing Aquatic Food Products. John Wiley and Sons.
New York.

Drying refers to the removal of moisture from a product. Although water may be removed by
mechanical application of pressure, adding salt, or using absorbent pads, evaporation of water
from the product surface is by far the most widely used drying technique. Dehydration, a term
often used interchangeably with drying, will be reserved in this discussion for use in referring to
the mechanical drying process, particularly those using a ground or minced product. Drying will
be used as a more general term meaning the removal of water from the product, usually by

Drying processes are used for preservation of aquatic products. Sometimes drying is the only
preservation process used, but it is also often used in combination with salting and/or smoking
for additional preservation.

Fish are typically 75 to 85 percent water. At atmospheric temperatures of 5°C or higher, fish
products are rapidly attacked by bacteria and molds, causing rapid quality deterioration. Since
mold and bacteria require water to support their life processes, drying fish products to a moisture
content below that required for survival by bacteria and molds greatly reduces the deterioration
rate and extends shelf life.

Drying Principles
The amount of water in a product can be expressed in several ways. The percent moisture (water)
content calculated on a wet basis (wb) is 100 times the weight of water in the product divided by
the total weight of the product (i.e., dry material plus water). Moisture content on a dry weight
basis (db) is 100 times the weight of water in the product divided by the weight of dry material in
the product. For products containing considerable water, the moisture content (M) on a dry
weight basis can exceed 100 percent. For example, calculate the moisture content of a 1000-gram
fish which contains 800 g of water and 200 g of dry matter.
                   800 g    
M   wb
          (100 ) 
                   1000 g     80 %
                            
                   800 g 
M   db             200 g   400 %
          (100 )        
                         
The moisture content of salted fish is sometimes quoted on a salt-free wet basis (Msfwb) or salt-
free dry basis (Msfdb). These moisture contents are calculated on the basis of the weight of the
fish or dry solids less the weight of salt in the product. The salt-free moisture content of a fish
with a total weight of 1000 g, of which 700 g is water, 50 g is salt, and 250 g is dry material
other than salt, can be calculated as follows:
                    700 g              
Msfwb  (100 ) 
                1000 g  50 g            73 . 6 %
                                       
                        700 g         
                1000 g  50 g  700 g   280 %
Msfdb  (100 )                        
                                      

If the weight of salt were not considered, the same fish would have the following moisture
                   700 g    
M   wb
          (100 ) 
                   1000 g     70 %
                            
                       700 g        
M   db
          (100 ) 
                   1000 g  700 g     233 %
                                    

Occasionally, data are also reported on a salt-free and fat-free basis using either a dry or wet
basis. Equations below show how the salt- and fat-free moisture content on a dry and wet basis,
respectively, are calculated.
                            weight of water in product                
Msffwb  (100 ) 
                 Weight of product  weight of salt  weight of fat   
                                                                      
                             weight of water in product                
                 weight of dry solids  weight of salt  weight of fat 
Msffdb  (100 )                                                        
                                                                       

Fat-free basis is sometimes used when the moisture content of a lean and a fat fish are compared.
Typically, fresh lean fish have a moisture content (wb) around 80 percent. In fresh fatty fish, the
sum of the moisture content (wb) plus the fat content generally is in the area of 80 percent. Thus,
fat tends to reduce the moisture content. For example, sardines with a 20 percent fat content
generally have a moisture content (wb) in the range of 60 percent (Waterman, 1976).

Changes in the amount of water in a product often are expressed as percent of water lost during a
drying process. The percent water lost is 100 times the ratio of the initial weight of water in the
product less the final weight of water in the product, divided by the initial weight of water in the
product. If a fish initially weighed 1000 g, of which 800 g was water and 200 g was dry solids,
and the fish was dried to a final total weight of 300 g, of which 100 g was water and 200 g was
dry solids, the percent water loss (PWL) would be
              800 g  100 g 
PWL  (100 ) 
                              87 . 5 %
                  800 g     
The product yield (PY) is 100 times the final total product weight divided by the original total
product weight. For the example above, the initial weight was 1000 g and the final weight was
300g. The yield is then
                300 g       
yield  (100 ) 
                1000 g        30 %
                            
Yield is often used when calculating the cost-benefit ratio of drying or when calculating
transportation costs of dried product from wet product purchased.


Moisture content is a value easily determined by oven drying and other methods. It is not,
however, a direct measure of the availability of water for bacteria and mold growth or for
enzymatic activity. Since these processes are the primary causes of product quality deterioration,

a direct measure of their ability to exist is helpful. Water activity (aw) defined as the ratio of the
vapor pressure in the product to that above pure water, is a better measure of the extent of
binding of water or the availability of water to support biological activity. Pure water is assigned
an aw value of 1. Fresh fish have an aw of above 0.95. Although the aw limit below which a
microorganism will not grow is species specific, as a general rule most spoilage bacteria will not
grow at an aw of 0.90 or below, most molds are inhibited below 0.8, and most halophilic bacteria
cease to grow below an aw of 0.75 (Waterman, 1976). Most spoilage bacteria cease growth below
25 percent Mwb, and molds rarely grow below 15 percent Mwb. The presence of salt causes
normal spoilage molds and bacteria to cease growth at higher Mwb typically 35 to 45 percent. For
additional details on the effects of salt on microorganism growth, see the discussion on salting.


Generalized mathematical models of drying in fish and other foods are available in the literature
(Jason, 1958; Charm, 1971). The solution to the diffusion and heat transfer equations involved in
drying involve differential and or integral calculus. Thus, detailed treatment of the mathematics
of drying are beyond the scope of this text. The general principles of drying will be presented,
and readers interested in a rigorous mathematical treatment of drying are referred to the available
references on drying (Jason, 1958; Charm, 1971).

Drying of fish is generally characterized by two types of drying. Initially, drying is governed by
evaporation from the surface or near surface areas. This period is referred to as the constant rate
drying period, because drying continues at a constant rate equal to the rate of evaporation from a
free water surface (Jason. 1958). Drying rate is thus a function of air velocity, air humidity,
temperature, product surface area, the amount of heat transferred from air to product per unit of
time, and other variables. Heat is required to evaporate water (2.258 kJ g at atmospheric
pressure). Thus, the drying rate is also a function of heat flow to the product.

As the surface water evaporates from the fish, water must be transferred from the muscle interior
to the surface before it can evaporate. This transfer takes place largely by diffusion of liquid or
gaseous water. Since diffusion is generally slower than surface evaporation, diffusion limits the
drying rate. Diffusion rate is a function of product structure, temperature, diffusion path length,
and other variables. This period is generally characterized by a slowly decreasing rate of drying,
at least partially due to the fact that the dryer product the further the water must diffuse to reach
the surface. Thus, this second period is referred to as the falling rate drying period.

Constant Rate Drying Period. The variables influencing drying during the constant rate are
those which effect evaporation from any free-water surface.
Air Velocity. Theoretically, increased air velocity results in increased drying rate. Practically,
this is true only within certain limits. During evaporation, water molecules must transfer through
an essentially stagnant surface film immediately above the water by diffusion. Once through the
surface film, they are transferred into the air largely by convective currents. The thickness of the
surface film is inversely proportional to air velocity over the surface. Since diffusion through the
surface film is often the rate limiting process in evaporation, higher air velocities result in a
higher evaporation rate. However, at high velocities, a unit increase in velocity produces much
less change in film thickness than does a unit change in velocity at low velocities. Increasing

velocity thus produces diminishing returns. In mechanical dryers, power requirements are
approximately proportional to the cube of air velocity. From an economic standpoint, increased
air velocity quickly becomes expensive. Practical guidelines indicate that air velocities in
mechanical dryers in the range of 1 to 2 m/s are economically feasible and provide a reasonable
drying rate. Higher velocities also tend to shorten the constant drying rate period (Waterman,
1976). Since the total drying cycle generally will be about the same length, shortening the
constant rate period is self-defeating.

Surface Area. Drying area is, for practical purposes, directly proportional to area exposed to
drying air. The more surface area exposed, the greater surface available for evaporation to occur.
It is also possible, particularly in mechanical dryers with air flow in one end and out the other, to
saturate the drying air before it reaches the last fish in the downstream position. When this
occurs, adding more fish surface area will not increase the drying rate. The problem under such
circumstances is poor dryer design or improper dryer operation.

Air Humidity. Humidity influences
drying rate by limiting the amount
of water the air can absorb. It also
reduces the driving force (i.e. the
difference in vapor pressure between
that at the fish surface and in the air)
which also slows drying.

It should be noted that too rapid
drying or drying with too low a
humidity results in a poorer quality
product. It toughens the fish surface
and can lead to reduced drying rates
during the falling rate period,
because of increased diffusion
resistance by the hard surface layer.
The effects of low humidity air on
salt fish during drying have also
been noted in the discussion of fish

Temperature. The effects of
temperature on drying rate result
from the effect of temperature on
heat transfer and effect of
temperature on relative humidity.
Warm air holds more moisture than
cold air. If air temperature is increased without the addition of water, the relative humidity drops.
Lower relative humidities favor more rapid evaporation and higher drying rates.

The amount of heat transferred to the product is proportional to the difference in temperature

between the air (or in the case of freeze drying, the plate temperature) and the product. High
temperature differences increase heat transfer and drying rate. High air temperatures can also
result in partial cooking of the product.

Product Thickness. Drying rates decrease with increased product thickness. A greater percentage
of the water present is removed during the constant rate period for thinner product pieces than for
thick ones, primarily because of the increased surface area to volume ratio for thinner pieces.
Because evaporation rates are higher during the constant rate period than the falling rate period,
thinner product pieces dry more rapidly. The diffusion path length is also shorter in thin than
thick pieces.

Length of Constant Rate Period. The more rapid drying occurs, the shorter is the constant rate
period. Rapid drying generally increases overall drying time. An exception to this rule is the
drying of salt fish where several short high rate drying periods are alternated with periods of
press-piling to allow movement of water to the surface. This latter process is, however, quite
labor intensive.

Jason (1958) presented data to show the
effect of air velocity on the length of the
constant rate drying period (Figure 11.6).
Increased air velocity shortened the
constant rate period up to velocities of
about 400 cm s. Above 400 cm/ s, air
velocity made little difference on the
length of the constant rate drying period. It
should be noted that Jason's (1958) results
were based on laboratory drying of
relatively small cod pieces in small
numbers. Commercial dryers could react
somewhat differently owing to fish
volume and other factors.

Jason (1958) also detailed the effect of air
humidity (wet-bulb depression) on the
length of the constant rate drying period
(Figure 11.7). For the air velocities used,
366 cm / s, wet bulb depressions
exceeding about 12°C had little effect on the length of the constant rate drying period. Wet-bulb
depressions less than 12° C increased the length of the constant rate drying period.

Salt Content. During drying of salt fish, particularly more heavily salted fish, a salt crust forms at
the fish surface. This crust is composed of approximately 80 percent salt, 10 percent protein, and
10 percent water (Linton and Wood, 1945). However, the specific characteristics of this salt crust
depend on the salt content of the fish and the drying conditions. When the drying potential is
high (i.e., high wet-bulb depression or low humidity) the salt crust has low permeability to liquid
water. Thus, evaporation occurs below the crust and water diffuses through the crust as vapor. As

the drying potential increases, the salt crust becomes tighter, reducing the diffusion rate and
hence the drying rate (Linton and Wood, 1945).

The salt crust's permeability rises with increasing relative humidity. Hence, at high relative
humidities salt fish act much like fresh fish in terms of drying. Press-piling of salt fish between
short drying periods allows water in the fish to migrate to the surface, rewetting the salt layer,
which reduces crust resistance to vapor flow. Thus, after press-piling the drying rate is higher
until the salt crust dries (Linton and Wood, 1945; Del Valle and Nickerson, 1968).

Drying rate in salt fish is thus determined to a large degree by the drying air relative humidity.
Air velocities above about 100 cm/s have little effect on drying rate of salt fish because of the
salt crust effects. Thus, relative humidities in the range of 45 to 60 percent and air velocities of
about 125 cm/s are generally recommended for salted fish. The salt crust permeability appears to
be optimal at relative humidities of 45 to 60 percent (Linton and Wood, 1945).

Falling Rate Drying Period. Drying during the falling rate period depends on the transport rate
of water from within the product to the surface rather than the drying potential of the air. Jason
(1958) showed that the falling rate drying period is characterized by two rate (diffusion)
constants, D1 and D2. The drying rate during the first falling rate period decreases with time, is
significantly lower than during the constant rate period, and is higher than during the second
phase of the falling rate period. The second drying rate constant D2 only applies after extended
drying times. The weight loss during this period is so low that it is difficult to get a reliable value
for D2 (Jason, 1958).
Drying during the falling rate period can be described by Fourier's equation
C            C              C              C
              2               2               2

       Dx             Dy             Dz
 t          x              y              z
                  2               2               2

where C is concentration; t is time; and Dx, Dy, and Dz are diffusion coefficients in the x. y, and z
directions, respectively.

Jason (1965) solved Fourier's equation for the case of a rectangular solid. Because the series
solution converges rapidly, little error is introduced by using only the first term of the series.
Jason (1965) then substituted into the solution and developed Equation 11.4 by showing that fish
flesh is essentially isotropic (i.e. Dx=Dy=Dz).
Wt  We         8       2  1    1   1  
               2  exp   D  2  2  2 t 
Wo  We                4   a   b   c  
Wt = Weight of material at time t.
We = Weight of material at equilibrium.
Wo = Initial weight of material.
Dx = Diffusion coefficient in the x direction (parallel to fish backbone).
Dy = Diffusion coefficient in the y direction (from top to bottom of fish's body).
Dz = Diffusion coefficient in the direction (from one side of fish to other side).
2a = Length of meat piece in the x direction.
2b = Length of meat piece in the y direction.
2c = Length of meat piece in the a direction.
t = Time.

The exponential term (2/4) D (1/a2 + 1/b2 + 1/c2)t is often referred to as the drying rate constant
. The form of  varies with the shape of the product. Equation 11.2 shows the solution for a
rectangular solid of dimensions 2a x 2b x 2c. For a finite cylinder of radius a and length L,  =
(4D/2)( 1/ a2 + 1/ L2).

If the logarithm of Wt -We versus time is plotted (Figure 11.8) two straight lines result. The two
lines correspond to the two different diffusion coefficients, D1 and D2, which exist during the
falling rate drying period. Typically the change in slope occurs at about 0.1 g water per gram of
non-aqueous material, and is associated with the uncovering of the uni-molecular layer of water
covering the protein molecules (Jason, 1965). In practical applications D2 rarely is observed
because moisture contents often are not reduced enough for D2 to be the governing diffusion

                                            Jason (1958) developed values for D1 and D2for
                                            several species of fish at 30oC (Table 11-3). He also
                                            indicated that both D1 and D2 were similar for all
                                            species of non-fatty fish and that increased fat content
                                            decreased the diffusion coefficient. Based on this
                                            conclusion Jason (1965) gave average diffusion
                                            coefficients for non-fatty fish as a function of
                                            temperature (Figure 11-9).

                                            Product shape. The size, shape and thickness of a
                                            product influence the drying rate as shown by
                                            equation 11-4. Diffusion is a function of thickness;
                                            hence thicker products require longer drying times.

                                            Air velocity. Diffusion from within the product to the
                                            surface is not influenced by air velocity. Drying
                                            during the falling rate period is thus independent of
                                            air velocity above some minimum value.

                                            Temperature. The diffusion coefficients D1 and D2
                                            increase with temperature. Drying will therefore
                                            proceed more rapidly with temperature increases at
                                            least until cooking or other product changes occur.

Relative humidity. Drying rate is a function of the difference between product initial and
equilibrium water content. Because equilibrium water content depends on relative humidity
drying rate will also change with relative humidity. However, this effect will be negligible except
when relative humidity is very high, because at the higher values the difference between the
initial and the equilibrium water content changes rapidly with relative humidity (Jason, 1958).

Salt content. Del Valle and Nickerson (1968) studied the drying of salted fish. They found that
salted fish exhibit two diffusion coefficients during the falling rate period. The magnitude of the

first falling rate diffusion coefficient, D1, depends on the salt content of the fish. The diffusion
generally increases, reaches a maximum, then decreases as salt content increases. Using pieces of
swordfish (Xiphias gladius) they found the maximum diffusion coefficient occurs at a salt
concentration of about 1 to 1.5 moles/L. Higher salt concentrations remove more water by
osmotic effects. Thus evaporative drying generally requires less drying time for heavily salted

Energy Use
Evaporating 1 g of water requires 2.258 kJ of energy. Thus removal of water by evaporation
requires considerable energy. Use of free solar energy is economically desirable but weather
conditions preclude use of direct sun drying of fish in many areas. Although drying air may be
heated by solar collectors in most areas economic feasibility has not been determined for most
locations. Baird et al. (1981) described a fish drier using solar collectors which was used in
Florida. Mechanical driers using fossil fuels have been widely used in the past. As fuel costs
increase, the economic advantage of mechanical driers will decrease.

TABLE 11.3 Values for several falling rate drying diffusion coefficients D1 and D2 for several
species of fish.
                                               Fat content   D1               D2
                                               (% of Wet
 Species                                       Weight)     (10-6 cm2/s    (10-6 cm2/ s)
 Catfish (Anarrhichas lupus)a                       0.102            3.61        0.795
 Cod (Gadus callarius)                               0.05 3.40 ± 0.366    0.81 ± 0.04c
 Conger eel (Conger conger)                         3.766            2.28        0.425
 Dab (Pleuronectes limanda)                          0.46            2.94         0.52
 Dogfish (Acanthius vulgaris)                          8.6           0.83         0.15
 Haddock (Gadus aeglefinus)                         0.105            3.25           0.6
 Halibut (Hippoglossus vulgaris)                    0.208            2.49         0.58
 Lemon sole (Pleuronectes microcephalus)            0.094            2.63           0.4
 Ling (Molva molva)                                 0.047            3.57         0.54
 Mackerel (Scomber scombrus)                        0.694            2.21         0.35
 Monkfish (Lophius piscatorius)                     0.094            3.06           0.4
 Saithe (Gadus virens)                              0.111            3.06           0.3
 Skate (Rata balls)                                 0.139            3.28         1.03
 Whiting (Gadus merlangus)                          0.036            2.72         0.48

Source. Jason, 1958.
  The common and scientific names for the species are those according to the Int. Fish Journal.
  Mean and standard deviation of 53 determinations.
  Mean and standard deviation of 9 determinations.

Drying Equipment

Drying of whole fish is most
often done using sun drying
or mechanical tunnel driers.
In many areas, fish are
simply laid on the beach or
rocks in the sun and allowed
to dry. This technique is
inexpensive but results in
considerable loss due to
spoilage, contamination,
rodents and insects. Figure
11-10 shows drying racks
used for fish. The racks are
generally placed about 1 m
above the ground, and the
product to be dried is placed
on the rack. Racks made of
wooden slats, rope mesh, or
vinyl-coated wire are widely
used. Racks allow air to
circulate around the product
and increase drying rate.
Since the product is above
ground, less is lost to rodents and insects.

Mechanical air dryers are the most widely used air dryers for fish. Mechanical dryers range in
sophistication from a simple enclosure with a fan to total air temperature and humidity control.
The totally controlled type dryer is depicted in Figure 11.11. Air humidity and temperature
control are maintained by cooling and/or heating the inlet air to achieve the desired conditions.
Under some conditions a certain percentage of the air may be recycled. Once conditioned, air is
drawn over the fish, which are hung on racks or placed in wire mesh trays. Air supplies heat to
evaporate the water and carry it away from the product. The exhausted air removes the water
vapor from the tunnel dryer.

Peters and Newton (1980) described a tunnel dryer used for drying salted fish. The dryer, 12 m
long by 1.6 m wide by 1.8 m high, provided for control of both temperature and humidity.
Normal load for this batch dryer was 4000 kg. Air velocity was variable up to 4.5 m/s. With a
full load, the dryer reduced the lightly salted fish to a M/ Mo ratio of 0.6 in 24 hours, where M =
moisture content (db) at the time in question and Mo is the moisture content (db) at the start of
the drying cycle. Energy used by this dryer (including fans and heating and cooling) averaged
0.7 to 1.0 kWh per kg of water evaporated when operating with a full load. Partial loading
increased the energy consumption per kilogram of water evaporated.

Although roller, spray, vacuum,
and other types of dryers are used
for drying various aquatic
products, these dryers are not
generally used for whole, dressed,
or filleted fish, shellfish, and
crustaceans. The operating
principles of these dryers are
described in most food science or
food engineering texts and will
thus not be discussed here.

Freeze drying has been used for
drying fish and other aquatic products. However, it suffers from difficulty in getting good surface
contact between the plate and the fish and from high cost. Poor plate-fish contact results from the
shape of the fish. Because of these problems, freeze drying is used only for special applications
where market conditions will support the additional cost.

Drying rate can be increased
and drying costs reduced by
use of the "accelerated freeze
dryer." This system is
essentially a freeze dryer, but
the plates are made of
expanded aluminum. This
innovation allows effective
heat transferred to the product,
such as fish fillets, while
allowing escape of water
evaporated from the plate side
of the product. Both these
advantages increase drying rates (Janson, 1965).


Shared By: